This proposal will investigate the immunologic mechanisms of tolerization to self antigen in humans with insulin dependent diabetes mellitus (IDDM). A theoretical goal of antigen specific immune therapy is the induction of autoreactive T cells secreting the regulatory cytokines TGFbeta-1, IL-4, and IL-10 that migrate to the organ and downregulate the disease process. We hypothesize that cytokine secretion of autoreactive T cells will be altered in subjects with IDDM receiving oral or parenteral insulin. These studies will be part of the multicenter Diabetes Prevention Trials (DPT-1) to investigate the clinical effects of parenteral and oral insulin administration, The investigations in IDDM contrast to that of multiple sclerosis patients orally tolerized with myelin as myelin basic protein and proteolipid protein represent sequestered brain antigens as compared to insulin, which is present in the serum. Thus the contrast of mechanisms associated with oral tolerance to these two antigens is of particular interest. We will first examine whether parenteral or oral administration of insulin change the frequency of insulin or GAD autoantigen reactive T cells as measured by proliferation, We will measure the frequency of insulin and glutamic acid decarboxylase (GAD) reactive T cells by culturing with antigen before and after parenteral or oral administration of insulin in the DPT-1 study group patients. Changes in precursor frequencies will be compared to tetanus toxoid (TT). Secondly, we will determine if parenteral or oral administration of insulin induce changes in the secretion of IL-4, IL-10, gamma-IFN or TGFbeta-1 of insulin of GAD reactive T cells. We postulate that insulin reactive T cells in subjects receiving insulin will secrete IL-4, IL-10, or TGFbeta-1 and thus have the potential to downregulate immune responses. We will measure antigen specific cytokine profiles to insulin, and control antigens GAD and TT. T cells at limiting dilution will be stimulated by antigen and cytokine secretion, intracytoplasmic staining and mRNA measured in short term T cell lines using a split well assay. Third, as high titers of insulin autoantibody (IAA), islet cell cytoplasmic autoantibody (ICA), or GAD auto-antibody can be predictive of IDDM and we have shown changes in IAA levels parenteral insulin administration, we will examine whether autoantibodies change with parenteral or oral administration of insulin. We hypothesize that the repertoire and quantity of autoantibodies will be altered with orally or parenterally administered insulin in linkage to changes in cytokine secretion of insulin reactive T cells. Autoantibody titers to insulin, GAD, and ICA will be compared to the frequency of IL-4, TGFbeta-1 and gamma-IFN secreting T cells and to the HLA-DQ type of the individual subjects. If the repertoire and pattern of autoantibodies are altered, they would offer a more efficient method of following immunologic changes in IDDM. Lastly, we will examine whether IL-4/TGFbeta-1 secreting T cells are derived from the same parent T cells as gamma-IFN secreting T cells. This will involve determining whether the epitope recognition of insulin reactive T cells changes with tolerance induction. If the same epitope is recognized, we will sequence TCR chains to see if they were derived from the same precursor T cell, In total, the results of these investigations may allow us to understand mechanisms of tolerance induction in subjects with autoimmune diseases.